Release of Nitrogenous Volatile Species from South African

Article pubs.acs.org/EF

Cite This: Energy Fuels XXXX, XXX, XXX−XXX

Release of Nitrogenous Volatile Species from South African Bituminous Coals during Pyrolysis Zebron Phiri,*,† Raymond C. Everson,† Hein W. J. P. Neomagus,† André D. Engelbrecht,‡ Barry J. Wood,§ and Bonny Nyangwa∥ †

Coal Research Group, School of Chemical and Minerals Engineering, North-West University, Potchefstroom 2520, South Africa CSIR Materials Science and Manufacturing, PO Box, 395, Pretoria 0001, South Africa § Centre for Microscopy and Microanalysis, The University of Queensland, St. Lucia QLD 4072, Australia ∥ Eskom Research and Innovation Centre, Lower Germiston Rd., Rosherville, Johannesburg 2095, South Africa ‡

ABSTRACT: The influence of typical South African coal attributes on the release of nitrogen into the volatile stream during pyrolysis was studied by utilizing three bituminous coals. The majority of South African coals are characterized by high mineral matter and are rich in inertinite maceral. Pyrolysis was conducted in a bench-scale fluidized bed (FB) at 740−980 °C, and also in a drop-tube furnace (DTF) at 1000−1400 °C. Levels of nitrogenous species in the volatile stream in the form of NH3, HCN, and tar-N were determined. Nitrogen functional forms of tars released at low temperatures were predominantly distinguished by high levels of pyrrolic nitrogen, followed by pyridinic and quaternary nitrogen, respectively. Tars liberated at 740 °C possessed similar nitrogen functional form attributes as those of parent coals. However, an increase in pyrolysis temperature caused a gradual increase in quaternary nitrogen as well as a concurrent decrease in pyrrolic nitrogen and a concomitant subtle decrease in pyridinic nitrogen. The analysis of nitrogen in tars was only confined to tars extracted from the FB. Vitrinite-rich and/or high mineral matter coal released high yields of nitrogenous species into the volatile stream at low FB temperatures. A large amount of NH3 was released relative to HCN under FB pyrolysis conditions. However, more HCN was released than NH3 during DTF pyrolysis. Two coals, one characterized by high mineral matter and being rich in vitrinite, and the other distinguished by relatively low mineral matter and being rich in inertinite, behaved similarly by reaching respective peak amounts of NH3 yields at 820 °C under FB pyrolysis conditions. On the contrary, an opposite profile displaying a slump at 820 °C was observed for HCN yields from the two respective coals. The third coal, a high mineral matter and inertinite-rich coal, released high NH3 yields and simultaneously the least HCN yields at 740 °C. Under DTF experimental conditions, both NH3 and HCN steadily increased with temperature in all coals. The low mineral matter and inertinite-rich coal released high yields of total volatile-N from 1000 to 1270 °C, only to be surpassed by the vitrinite-rich/high mineral matter coal at 1400 °C. The inertinite-rich/high mineral matter coal released the least throughout the entire DTF temperature range. The total mineral matter content of the coals played a significant role toward the nitrogen product distribution. On the other hand, the total reactive macerals also influenced the emission of volatile species at 1130−1400 °C DTF temperature range. The yields and composition of the released nitrogenous species have been attributed to a combination of mineral matter content, petrographic properties of the parent coals, and the utilized conditions. Pyrolysis temperature, coal particle size, and residence time also play a significant role toward the yields and composition of the released nitrogenous species.

■

INTRODUCTION

necessary to comply with the ever increasing stringent legislation on nitrogen oxides emission from coal combustion. South Africa’s power generation industry is heavily dependent on coal combustion, hence the need to be more vigilant and heed the call to significantly reduce emissions. The partitioning and product distribution of coal nitrogen between volatile species and char nitrogen have been perceived to be influenced by coal properties, encompassing parent coal composition, attributes of the respective char residue, and the surface characteristics.22,23 Nitrogen is preferentially retained in char at low temperature; however, a large fraction is released into the volatile stream at high temperatures.24−31 In addition to pyrolysis temperature, other conditions which include heating

Introducing coal into a combustion unit initially gives rise to devolatilization (pyrolysis) causing the partitioning of coal-N into volatile-N and char-N.1 The volatile-N stream constitutes of tar-N, HCN, NH3, and N2.2 HCN and NH3 have been identified as the main precursors of oxides of nitrogen during coal combustion in several studies.3−13 The volatile-N species influence the respective final oxides of nitrogen pollutants produced during the subsequent combustion. Nitric oxide (NO) and nitrogen dioxide (NO2), often designated as NOx, are potent pollutants that have been attributed to causing acid rain, ground level ozone, and photochemical smog. Another oxide of nitrogen, nitrous oxide (N2O), is a potent greenhouse gas, and it has been blamed for indirectly depleting the ozone layer.14 Numerous studies have been conducted with the quest to comprehend the chemistry behind the formation of tar-N, HCN, NH3, and char-N as precursors of NOx and N2O.13−21 However, further work is still © XXXX American Chemical Society

Received: October 31, 2017 Revised: March 2, 2018 Published: March 5, 2018 A

DOI: 10.1021/acs.energyfuels.7b03356 Energy Fuels XXXX, XXX, XXX−XXX

Article

Energy & Fuels

fuel nitrogen in coal, whether it is infused in aromatic groups or part of the peripheral heterocyclic structures, determines the form of nitrogen species released during devolatilization, implying that the nitrogen species emitted during devolatilization having been correlated to the coal nitrogen functionality. Hoogendoorn et al.61 proposed an overview of different paths of fuel nitrogen by suggesting that N-5 and N-6 nitrogen are liberated as HCN in residual chars, while the amino groups in raw coal are evolved as NH3. Pyrrolic nitrogen is susceptible to decomposition during heat treatment;26 the process is perceived to occur through degradation (ring expansion), of which the 5-membered ring embodying pyrrolic nitrogen opens up due to high temperature.46,62,63 However, some researchers,55,56,64 have questioned and disputed the correlation of fuel nitrogen functionality with the liberation of nitrogenous species during pyrolysis and/or combustion. Most coals in Southern Africa are typical of the Gondwana region (southern hemisphere) coals, they are characterized by high levels of mineral matter and inertinite content.65−67 Southern hemisphere coals vastly differ from the Laurasian region (northern hemisphere) coals that are typically high in vitrinite and low mineral matter. Mineral matter inherently exists in coals in the form of alkali, alkaline earth, and transition metal cations.68−73 Coals with high mineral matter content constitute of appreciable quantities of Na, K, Mg, Ca, and Fe cations in ion-exchanged forms.74 The presence of some of the mineral species may alter the partitioning of nitrogen into the volatile stream and char during devolatilization.2,21,70,75,76 Yan et al.75 reported that pyrolysis of demineralized coals tends to suppress the release of nitrogen volatile species. Fe, Ca, K, Si, and Al increase the conversion of coal-N to NH3, whereas Na promotes the formation of HCN.75 Mori et al.21 pointed out that, at 900 °C, the presence of Fe promotes the formation of N2; however, it also hinders the conversion of nitrogen to HCN and NH3. Ohtsuka and co-workers69 demonstrated that the influence of Na, K, and Ca on the composition of HCN, NH3, and N2 product species varies to a great extent with temperature from 450 to 950 °C. The particular research group,2,21,51,69,71,77−79 focused on the formation of N2 during coal pyrolysis, concluded that char-N is the major source of N2, and also showed that the conversion of char-N to N2 is catalyzed by Fe and Ca cations. There is limited information in the open literature with regard to the release of nitrogen species from inertinite-rich coals. Inertinite-rich coals generally display higher aromaticity than the vitrinite-rich counterparts.34,80 However, vitrinite-rich coals tend to have a higher degree of hydrogen content, H/C atomic ratio, aliphatic moiety as well as aliphatic side chains than coals that are rich in inertinites.34,81 Zhao et al.81 pyrolyzed vitrinite and inertinite maceral fractions and reported that the inertinite produced lower tar and gas yields, as well as possessing higher thermal stability. Given and co-workers82 found that N in maceral concentrates from a number of British coals were as follows: vitrinite (1.6−2.0% N, dmmf) > exinite/liptinite (1.1−1.4%) > inertinite (0.3−1.2%). Rajan and Raghavan83 reported that a coal that is rich in liptinite released moderately higher levels of nitrogen as compared to a vitrinite-rich coal, in spite of the vitrinite-rich coal containing slightly higher nitrogen. They ascribed the phenomenon to fuel nitrogen breaking free more easily from the liptiniterich coal particles than in high-vitrinite-rich coals. Understanding the chemical and physical processes that determine formation of precursors of nitrogen oxides during devolatilization remains a challenge even though much research has been conducted in the last four decades. South African (SA)

rate, residence time, and pressure also influenced the formation and distribution of NH3 and HCN. Tar is the primary means of nitrogen release from coal.20 Essentially, almost all the nitrogen in tars liberated during coal pyrolysis exists in the form of heterocyclic aromatic structures, and the portion of light volatile-N gases is quite negligible during initial phases of pyrolysis.15,20,29,31 Tar may become the significant source of nitrogen production into the volatile stream as the secondary pyrolysis proceeds.18,32 Char decomposition during pyrolysis occurs at high temperatures and considerably longer residence times than during the primary devolatilization phase both lead to the liberation of additional nitrogen species into the volatile stream, mostly in the form of HCN.14,16,28,29,33 High temperatures may result in the disintegration of remaining heteroatomic aromatic rings of char, which constitutes mostly of comparatively stable quaternary nitrogen,34 thereby releasing more nitrogenous species into the volatile stream. The release of HCN has been found to be prevalent during rapid pyrolysis experiments, while NH3 has been reported to be dominant under low heating rate conditions.5,23,35 The HCN/NH3 ratio has been reported to increase with higher heating rates in both coals and biomass pyrolysis.5,17,36 It is widely reported that HCN is the major product of bituminous coals devolatilization, while NH3 is mainly released from low rank coals and biomass.5,16,35,37−40 Considerable research has been done on the liberation of nitrogenous species during coal pyrolysis.3−11,21,41−43 However, some researchers have opted to work with model compounds to avoid the experimental constraints resulting from the complex heterogeneity nature of coal.12,13,44−50 Model compounds are bound to encounter a drawback due to incorporated solitude species which may fall short or exacerbate the reaction in forecast as compared to the synergistic effect which may arise in coal pyrolysis as a result of the combined effect of N neighboring heterogeneous components which include minerals, functional groups, and free radicals.51 However, model compounds have shed some light on probable mechanisms of N release during pyrolysis. Numerous mechanisms thought to lead to the formation of NH3 and HCN have been proposed.7,8,25,37,52−57 It has been suggested that HCN is initially formed and NH3 is subsequently generated through direct hydrogenation of HCN and N-sites by hydrogen generated in situ when the solid fuel undergoes pyrolysis.8,52,53 Bassilakis et al.17 suggested that NH3 is principally generated from the heterogeneous reactions between char and HCN. Yuan and co-workers58 proposed that NH3 is formed directly during the primary stage of rapid pyrolysis through cracking of nitrogen functional forms. The reaction may proceed via ·N, ·NH, and ·NH2 free radicals being released from the fuel and then merging with ·H or H2 under the influence of high energy impact to form NH3. According to Tan et al.,6−8 the ·H radicals that are liberated during pyrolysis are adsorbed on the char surface, subsequently attacking the heterocyclic nitrogen bonds leading to the formation of NH3. Through molecular modeling, Espinal et al.59 used density functional theory to illustrate the dissociation of the C−N bond releasing ·NH2 radicals to the gas phase and the subsequent formation of NH3, through either homogeneous or heterogeneous hydrogen abstraction or recombination reactions. However, Kambara and co-workers25,54 suggested that the two components are formed simultaneously, and also highlighted that the distribution of nitrogen species during coal pyrolysis and combustion can be quantitatively predicted based on the nitrogen functionality in the substrate coals determined from the XPS N 1s measurements. Kelemen and co-workers60 pointed out that the nature of B


Article

Energy & Fuels Table 1. Properties of the Utilized Coalsf ultimate analysis (wt %, dmmf)d

proximate analysis (wt %, d.b.) coal name

MM %

Glisa Lethabo Matimba

22.0 43.3 36.6

a

VM %

b

c

maceral analysis (v/v %, mmb)e

FC %

C

H

N

S

O

vitrinite

liptinite

inertinite

56.4 38.5 39.9

82.8 79.5 83.6

4.6 4.2 5.3

1.9 1.9 1.8

1.0 1.1 2.1

9.6 13.2 7.1

18 15 50

3 3 3

69 46 23

21.6 18.2 23.5

a

Mineral matter (MM) analyses conducted according to ISO 1171:201087 and the Parr formula.88 bAnalyses conducted according to ISO 562:2010.89 cAnalyses conducted by difference. dAnalyses conducted according to ISO 29541:2010.90 ePetrographic analysis carried out using ISO 74042:1985 guidelines.91 fProximate results are reported on dry basis (d.b.), while elemental results are reported on dry, mineral matter free basis (dmmf). Maceral results are presented on mineral matter basis (mmb).

Table 2. Metal Compositions of Utilized Coal Samples metal composition (wt %, coal d.b.) coal

Al

Ca

Fe

K

Mg

Na

Si

Ti

Glisa Lethabo Matimba

2.68 5.52 4.50

2.12 1.60 1.30

0.79 1.12 2.25

0.15 0.24 0.25

0.54 0.30 0.30

0.02 0.04 0.02

2.60 8.90 9.50

0.16 0.33 0.27

utilized in FB experiments. The residence time of the gases in the FB ranged from 1.10 s, recorded at 980 °C, to 1.36 s at 740 °C. However, solid particles experienced a residence time of ±9.4 min in the reactor. In DTF experiments, fine coal particles of −75 μm were fed at an ≈1 g/min and pyrolyzed at 1000, 1130, 1270, and 1400 °C, respectively. The nitrogen gas flow rate ranged from 17.9 to 23.7 Nl/min so as to achieve a residence time of ≈1.8 s at each stipulated temperature throughout the DTF pyrolysis experimental program. Sampling and Analyses of N Volatile Species. The nitrogenous species reported in this study were sampled at the outlet of each specified reactor. Sampling tar entailed bubbling the flue gas through a multicomponent sampling train containing dichloromethane (DCM). The DCM dissolved the tar constituent of the gas emitted from the FB or DTF. Impingers containing DCM were placed in an ice bath at 4 °C to enhance condensation and trap the condensable organic species in the gas. The DCM in the collected samples was allowed to evaporate under controlled temperature conditions. This technique is usually employed in the analysis of coal tar pitch volatiles (CTPV). The majority of soluble and suspended tar components were left behind after DCM evaporation that was conducted under controlled conditions. The mass of tar was determined by weighing the solids left behind after total evaporation of the DCM sample. The tars were only successfully collected from FB pyrolysis experiments. No quantifiable amount of tar yields were obtained from the DTF operation due to the low coal feed injected into the reactor, and also as a result of tar condensation on the inner surface of the water cooled sample collection probe. Subjecting coal to high heating rates particularly at high temperatures tends to produce low tar yields.58,92,93 The speciation of nitrogen functional forms in tars that were collected at 740−980 °C were determined by X-ray photoelectron analysis (XPS). A Kratos Axis Ultra X-ray photoelectron spectrometer situated at The University of Queensland (Australia), operated at a base pressure of 1 × 10−8 Torr and a power of 225 W using monochromatic Al Kα radiation, was used to obtain XPS N 1s data. A detailed procedure on XPS N 1s spectra deconvolution and peak allocation is outlined elsewhere.24 The analyses of HCN and NH3 absorbed in respective solutions were conducted by Chemtech Laboratory Services (Pretoria, South Africa) using a Dionex DX-100 ion chromatograph equipped with a Dionex CD-20 conductivity detector. A multicomponent sampling train consisting of impingers containing 0.1 N NaOH was used to absorb the released HCN. The HCN that is present in the reactor outlet gas stream reacts with the NaOH to form CN− ions, which are retained in the alkaline solution until analyzed by ion chromatography in accordance to the U.S. EPA Method OTM 29.94 The CN− ions were separated in a Dionex AS-7 column and subsequently quantified. A similar sampling train setup was utilized for sampling NH3; however, an acidic solution of 0.1 N H2SO4 was placed in the impingers to absorb the NH3. The NH3 existing in the flue gas was converted to NH+4 ions by the H2SO4 solution

bituminous coals possess the typical attributes of southern hemisphere Gondwana coals, which differ from the northern hemisphere Laurasian coals in being variable between regions and seams. Most SA coals are generally characterized by high mineral matter content; the majority are rich in inertinite and basically have low calorific values.65,84 The primary goal of this study was to determine the influence of the typical SA coal attributes on the release of NH3, HCN, and tar nitrogen speciation. There is relatively limited information reported in the open literature on nitrogen oxides precursors and the distribution of tar nitrogen released from coals with high mineral matter and different petrographic attributes.

■

EXPERIMENTAL SECTION

Coal Samples. The three SA bituminous coals that were utilized in this study were also used by Phiri and co-workers.24,34 Prior to the pyrolysis experiments, the parent coals were subjected to the conventional proximate and ultimate analyses, as well as petrographic analysis. The respective chemical properties of the coals are given in Table 1. According to ISO 11760:2005 classification of coals,85 all three coal samples are classified as bituminous, medium rank C coals. The detailed petrographic analysis of the three coals is outlined elsewhere.34 The detailed petrographic analysis shows that Matimba coal contains 58% (vol, mmb) of total reactive macerals, followed by Glisa and Lethabo coals with 47% and 32% (vol, mmb), respectively. Glisa and Lethabo coals are both inertinite-rich; nonetheless, Glisa coal possesses substantial reactive semifusinite and reactive inertodetrinite, leading to a much higher portion of total reactive macerals. The mineral composition of the coal samples was analyzed by utilizing X-ray fluorescence (XRF) spectroscopy analysis according to the ASTM D4326-4 guidelines.86 The analysis was carried out on an ARL ADVANT’X instrument, of which the X-ray tube was operated at 2500 W (50 kV and 50 mA) to generate Rh Kα radiation. The coal samples were heated at 815 °C to obtain ashes based on ISO 1171:2010.87 Table 2 outline the concentrations of elements that incorporate Al, Ca, Fe, K, Mg, Na, Si, and Ti that are presented in wt % on coal dry basis. Pyrolysis Experiments. Pyrolysis experiments were conducted in a bench-scale fluidized bed reactor and drop-tube furnace in ultra-highpurity N2 (99.995%) obtained from Air Liquide (South Africa) at heating rates exceeding 104 °C/s. Both reactors are described in detail by Phiri and co-workers.24 In the fluidized bed experimental campaign, ±1 mm coal particle sizes of the respective coal samples were mixed thoroughly at a ratio of 1:1 with fluidizing material comprising 0.3−0.65 mm silica sand and fed into the reactor at 21 g/min; simultaneously N2 gas was introduced through the bottom of the reactor at 18.4 Nl/min. Four different temperatures in the range of 740−980 °C inclusive, were C


Article

Energy & Fuels

Figure 1. Volatile nitrogen released at different temperatures from FB and DTF pyrolysis of three South African coals.

in DTF were determined by difference.58,73 Char-N yields were the most abundant within the N product distribution from FB, with the exception of Matimba 900 and 980 °C, followed by N2, tar-N, NH3, and HCN, respectively. The emitted volatile species seem to be influenced by the pyrolysis conditions and the coal type. The inertinite-rich and relatively low mineral matter Glisa coal released less volatile-N species, whereas the high mineral matter and vitrinite-rich Matimba coal converted a substantial portion of coal-N under FB pyrolysis conditions. The char-N yields emanating from pyrolysis of Glisa and Lethabo coals in FB represented the dominant nitrogenous species. In essence, more fuel nitrogen is retained in the chars derived from inertinite-rich coals. The rate at which nitrogen was released was much slower than the rate of release of total volatiles during the relatively low FB temperature range.97 Matimba FB char-N yields ranged from about 40−50%, while N2 marked the dominant volatile-N component. Besides the N product distribution at 1000 and 1130 °C for Glisa and Lethabo coals, respectively, the determined tar-N + N2 exhibit prominence across the DTF pyrolysis temperature ranges for the three coals. Significant coal-N for Glisa and Matimba at 1130−1400 °C is partitioned into tar-N + N2. Coal pyrolysis at elevated temperatures and high heating rates usually release low tar yields,58,92,93 so the tar-N + N2 portion can be projected to be largely constituting N2. High temperatures may cause disintegration of the remaining heteroatomic aromatic rings of char-N, which constitutes mostly of comparatively stable quaternary nitrogen,34 and subsequently releasing more nitrogen species into the volatile stream. Quite a number of researchers51,70,71,77,98 attested that most of the N2 emanates from char-N and have also indicated that conversion of fuel-N to N2 increases appreciably with increasing pyrolysis temperature. This is one of the most likely reasons for the apparent decreased yields of char-N, HCN, and NH3 that are shown in Figure 2b. In addition, several other studies also reported that the presence of metal cations resulted in the reduction of HCN and a concomitant increase in the formation of N2 during pyrolysis.21,69,78,79 Char-N is the major source of N2 during high temperature pyrolysis, and the phenomenon is perceived to be catalyzed by Fe and Ca cations.2,21,51,69,71,77−79 Nitrogen Speciation in Tars. In general, the tar yields from all three coals displayed a gradual decrease as temperature increased. Nelson and co-workers92 reported a similar trend. The vitrinite-rich Matimba coal released more tar at each temperature across the entire FB temperature range, followed by the inertinite-rich Glisa and Lethabo coal, respectively. XPS N 1s analysis revealed that all the 3 parent coals primarily constituted of pyrrolic-N (N-5), followed by pyridinic-N (N-6) and quaternary-N (N-Q), respectively,24 as illustrated in Figure 3. The results conform to widely published literature on raw coals.46,60,99 The XPS N 1s spectra of coal tars liberated at 740 °C are also shown in Figure 3; the

and subsequently analyzed through ion chromatography following the U.S. EPA Method CTM 27 guidelines.95 A Dionex CS-12 column was used for separation prior to the quantification of NH+4 through conductivity detection. It is noteworthy that the NH3 yields may contain contributions from the hydrolysis of HNCO to form NH+4 ions; however, the yields of HNCO are sparingly low when compared to NH3.8,9,64,75,96 HCN and NH3 data were evaluated as an average of three analysis results, and the experimental error was determined at 95% confidence interval. The deduced experimental error ranged between 2.034% and 9.155% inclusive.

■

RESULTS AND DISCUSSION Nitrogen Distribution. Figure 1 illustrates a portion of coal-N that was released as volatile-N during FB and DTF pyrolysis. Volatile-N released from coal pyrolysis could be in the form of NH3, HCN, tar-N, or N2. The nitrogenous volatile species released from FB pyrolysis (Figure 1a) indicate that the conversion of vitrinite-rich Matimba coal-N was quite significant, releasing 47.7% N at 740 °C, and gradually increasing to 63.8% at 980 °C. However, the coal-N that migrated into the volatile stream during pyrolysis of inertinite-rich Glisa and Lethabo coals was constant, just below 20%, throughout the entire FB temperature range, with the exception of Lethabo coal that released 29.4% of its total nitrogen at 740 °C. Under FB pyrolysis conditions, the fraction of nitrogen released was dependent upon petrographic attributes, with the vitrinite-rich coal releasing a larger fraction of its nitrogen than inertinite-rich coals. A substantial portion of coal-N was converted into volatile-N during DTF pyrolysis (Figure 1b) of the three coals. Nitrogen release increased significantly with the pyrolysis temperature; Pohl and Sarofim28 reported similar findings. A significant fraction of coal-N being converted into volatile-N was observed as pyrolysis temperature increased. Matimba coal released more nitrogen into the volatile stream under DTF conditions; however, the difference to Glisa coal was quite negligible unlike to FB pyrolysis, implying that the influence of total reactive macerals is more pronounced at high DTF temperatures. Volatile-N from Lethabo coal lagged behind, even though it also displayed an upward trajectory with pyrolysis temperature. The nitrogen in vitrinite-rich coal particles is released more readily than in inertiniterich coals. Vitrinite macerals are generally richer in hydrogen than inertinite macerals. The increased hydrogen content and high volatile-matter content of the vitrinite-rich Matimba coal are reflected in Table 1. The inertinite-rich Glisa and Lethabo coals are more aromatic than the vitrinite-rich Matimba coal,34 implying that more nitrogen atoms in inertinite-rich coals are within aromatic structures, hence the increased difficulty of release. The yields of different nitrogenous species acquired from pyrolysis of the three coals in FB and DTF at various temperatures are illustrated in Figure 2. The N2 in FB, and the tar-N + N2 D


Article

Energy & Fuels

Figure 2. Yields of various forms of nitrogen emanating from FB and DTF pyrolysis of three South African coals at different temperatures.

Figure 3. XPS N 1s spectra depicting the nitrogen functional forms of the three raw coals and the respective tars released at 740 °C. The N 1s XPS spectrum of each tar is placed beneath that of the corresponding parent coal. E


Article

Energy & Fuels

Figure 4. XPS N 1s spectra illustrating the transformation of nitrogen functional forms in tars released from the three coals in a fluidized bed from 820−980 °C.

composition of nitrogen functional forms is essentially similar to that of the respective parent coals as deduced from the XPS N 1s spectra, implying that the tars released at the specified temperature are also predominantly composed of N-5, followed by N-6 and N-Q, respectively. During the first stage of devolatilization, a portion of coal-N is emitted rapidly with the tars, but with negligible or no recognizable selectivity with regard to nitrogen functionality,14 implying that, in the absence of secondary reactions, N-5 and N-6 species in coal are released intact as components of aromatic structures in tar. Solomon and Colket15 reported that tar possessed similar nitrogenous structural attributes as that of the respective parent coal through the comparison of 13C NMR and infrared spectra of parent coals and the respective coal tars. Utilization of the XPS technique, introduced years later, validated the observation and further elucidated the component functional forms through deconvolution of XPS N 1s spectra. The effect of temperature on nitrogen speciation of the released coal tars is illustrated in Figure 4. It is apparent that increasing pyrolysis temperature resulted in change in the nitrogen functional forms of tars as depicted by the change of sub-peaks on the curveresolved XPS N 1s spectral shapes. Heat treatment resulted in significant levels of edge located N-5 and N-6 in the parent coals being transposed into the volatile stream mainly as tar; hence the tars were dominated by nitrogen functional forms that resemble parent coals’ nitrogen speciation. Almost all the coal tars liberated from the three coals at 740−900 °C were fundamentally composed of N-5; nonetheless, a gradual increase in N-Q and simultaneous appreciable reduction of N-5 occurred as

pyrolysis temperature increased, such that the tars liberated at the highest temperature of the FB range were dominated by N-Q. The released coal tars subsequently undergo secondary cracking which generates light gases that constitute additional HCN and NH3.14,15 A subtle concomitant reduction of N-6 with increasing temperature was also observed. This is evident of secondary pyrolysis taking place, through the rupture of the heterocyclic aromatic structures that are housing pyrrolic- and pyridinic-N to release light nitrogenous species. However, Li et al.96 reported a different pattern that prescribed a relative increase in pyrrolic nitrogen in tars produced at 600−800 °C with respect to other forms of nitrogen in coal. Tars liberated from Matimba coal at 820−980 °C exhibited a fourth peak on the XPS N 1s spectra which is attributed to protonated and/or oxidized pyridinic nitrogen (N-X). The observation on the increasing proportion of N-Q with temperature in coal tars indicates that some of the N-Q species are sufficiently strong and intact to survive the devolatilization process or reform from their components. This entails that some of the quaternary nitrogen may be created during rapid heat up pyrolysis. The process proceeds through nitrogen substituting for carbon in condensed polynuclear aromatic structures, partially aromatic/heterocyclic systems giving rise to N being covalently bonded to three C atoms (N-Q). This implies that elevating pyrolysis temperature prompted the N situated on the boundaries of the graphene layers of tar molecules to become positioned on the inner parts of the graphene structures due to condensation and/or polymerization of polynuclear aromatic structures. Stańzyk et al.100 and Pels et al.46 suggested that F


Article

Energy & Fuels

Figure 5. NH3 and HCN expressed as fractional N release of coal-N during FB and DTF pyrolysis of three South African coals.

The results obtained indicate that the release of NH3 and HCN is influenced by a combination of coal properties and pyrolysis conditions. Under FB pyrolysis conditions, at 740 °C, the high mineral matter and inertinite-rich Lethabo coal released the most NH3. Figure 5a indicates that the release of NH3 of the three coals at 740 °C were in the order of Lethabo > Matimba > Glisa, which is also the order of mineral matter content of the coals as indicated in Table 1. Further, HCN released at 980 °C exhibited a similar order of yields, which is also the order of the mineral matter content of the parent coals. These observations suggest that the catalytic effect of inherent minerals could have had an influence on the release of NH3 at 740 °C and HCN at 980 °C.2,51 Therefore, the effect of mineral matter on the release of a particular nitrogenous volatile species is also dependent upon the pyrolysis temperature. Yan et al.75 pointed out that Fe, Ca, K, Si, and Al heighten the coal-N conversion to NH3 during pyrolysis, while Na promotes the formation of HCN. Fe and Ca have been reported to increase the decomposition of HCN to N2,1,21 concurrently reduce the conversion of fuel-N to HCN, while simultaneously promoting an increased release of N2.2,21,51,70 Wu and co-workers73 monitored the influence of Ca and Fe on the release of N2 during pyrolysis of model coals and determined that more N2 was released when the additions of Ca and Fe were increased from 0.5 to 3.0 wt %. The release of NH3 at 820 °C seemed to be influenced by the maceral composition of the coals. The Matimba coal which has high vitrinite content released the most NH3 at 820 °C, whereas the other two inertinite-rich coals, Glisa and Lethabo, both released much less amounts of NH3. Under DTF conditions, at 1000 °C, Glisa coal released the most NH3 and HCN. The order appeared to be influenced by the reverse of mineral matter content; Glisa coal

N-5 and N-6 amalgamated concurrently with the ring opening process, leading to the formation of stable N-Q. The nitrogen atom in N-Q is shared among two or three aromatic rings; the three bonds connecting to the N atom may break simultaneously when subjected to intense pyrolysis conditions, providing a high energy impact within a short period. This could lead to possible formation of N radicals preceding the formation of N2, NH3, and HCN. NH3 and HCN Release. Ammonia and hydrogen cyanide released from FB and DTF during pyrolysis of three South African bituminous coals are shown in Figure 5. The fraction of coal-N released as NH3 in FB at temperatures of 740−980 °C is illustrated in Figure 5a. NH3 emission from Lethabo coal displayed a maximal at 740 °C and a minimal at 980 °C, while Glisa and Matimba displayed their respective maximum NH3 yields at 820 °C, ranging from about 3.2−7.4%. The effect of temperature on the release of NH3 shown in Figure 5c for Glisa and Matimba coals was also observed by Tsubouchi and Ohtsuka2 on six Australian coals with a maximum NH3 release occurring at around 800−850 °C. This trend was also confirmed on coals investigated by Li and Tan.8 HCN emission principally increased with temperature as shown in Figure 5b; however, the HCN release from Glisa and Matimba decreased and displayed a minimum value at 820 °C, as opposed to the NH3 yields at the same temperature. Figure 5c,d shows the effect of temperature on the respective release of NH3 and HCN during DTF pyrolysis. In general, HCN was the predominant precursor of nitrogen oxides, while only small portions of coal-N were converted into NH3. The release of NH3 and HCN from Glisa coal was higher than that of the other coals at 1000 °C; however, at 1130−1400 °C, the release of NH3 was higher from Matimba coal. Similar proportional amounts of coal-N were released as NH3 at 1270 and 1400 °C for Glisa and Lethabo coals. G


Article

Energy & Fuels

Figure 6. Ratio of NH3 to HCN released during pyrolysis of the three coals in FB and DTF.

of different pyrolysis conditions used. FB pyrolysis released relatively more NH3 compared to HCN for all three coals. The FB pyrolysis experiments were characterized by high residence time, large coal particle sizes (±1 mm), and relatively low temperatures. The high yields of NH3 may be conceptualized as a result of HCN initially forming during pyrolysis.17 The formation of NH3 occurs by virtue of hydrogenation of HCN which predominantly occurs homogeneously as a result of the inherent H2 released during pyrolysis. Hydrogenation also takes place heterogeneously on the char surface. Other reactions involving free radical mechanism may also occur on the char surface to convert HCN to NH3.6,7 In addition to hydrogenation, the formation of NH3 is also perceived to emanate directly from the solid fuel matrix;17 however, the conditions favoring the hydrogenation of HCN on the char surface results in high yields of NH3. Carpenter and others14 stated that, despite the observed trends of HCN and NH3, it remains speculative whether HCN or NH3 is emitted discretely or whether, and to what extent, NH3 is an outcome of secondary reactions. Johnsson1 reviewed several studies that pointed out that the conversion of coal-N to HCN and NH3 was dependent on pyrolysis conditions, with a higher proportion of NH3 to HCN reported in slow heating-rate pyrolysis compared to rapid heating-rate pyrolysis. The DTF pyrolysis experiments were characterized by high temperatures and small particle sizes. The influence of coal properties on the formation of volatile-N species varies with the heating manner, which is the final temperature in this instance. The total yields of nitrogen in light volatile nitrogenous species as deduced from the summation of NH3 and HCN (NH3 + HCN) is shown in Figure 7. Since the yield of NH3 in FB was far greater than that of HCN, the trend of total volatile-N species almost resembles the yields pattern of NH3. The significant decrease of total volatile N species with temperature (from 740 °C for Lethabo, and from 820 °C for both Glisa and Matimba coals) in the FB is attributed to the increased conversion of tar and char nitrogen into gaseous N2. The total volatile-N yields emitted during pyrolysis from the three coals in the DTF (Figure 7b) increased with temperature. Glisa coal released more volatile-N from 1000 to 1270 °C; however, the yields obtained remained almost unchanged between 1270 and 1400 °C. Nonetheless, the total volatile-N yields from Matimba coal increased significantly at 1400 °C. This observation could be attributed to the low volatile content and the slightly higher aromaticity of the inertiniterich Lethabo coal. At 1400 °C, all the volatiles had been driven off,24,34 while the moderately high aromaticity suggests that the nitrogen atoms exist in the center of the clusters in the form of N-Q,24 which is more stable than the N-5 and N-6, hence

has the least and Lethabo coal has the most, leading to the sequence: Glisa > Matimba > Lethabo. However, the yields of NH3 and HCN at 1130−1400 °C gave the impression that they were influenced by the vitrinite maceral; Matimba coal released the most, while the inertinite-rich Glisa and Lethabo coals released approximately equal amounts. Therefore, the catalytic effect of the inherent minerals in parent coals seemed not have a major influence on the release of HCN and NH3 under the utilized DTF conditions. Tsubouchi101 showed that the yields of tar-N, HCN, and NH3 were almost unchanged when residence time was increased to 120 s regardless of the coal type; however, the N2 yields increased while the char-N decreased with increasing residence time. Tsubouchi101 further highlighted that small quantities of inherent Ca and Fe seem to enhance the formation of N2 from devolatilized char-N. In their earlier work, Tsubouchi and co-workers51,70,71 demonstrated that Ca2+ ions doped in low rank coals promoted the formation of N2 from char-N regardless of the heating rate when coals are pyrolyzed at temperatures exceeding 1000 °C. Table 2 shows that the three coals possessed substantial amounts of Ca which ranged between 1.30% and 2.12%, as well as 0.79−2.25% Fe. Low mineral matter content seemed to favor the liberation of light volatile-N species under the given DTF conditions. Tsubouchi and Ohtsuka2 reported that slow heating rate pyrolysis of acid demineralized low-rank coals at 1000 °C in a fixed bed reactor increased HCN and decreased NH3 output. This implies that low mineral matter or absence of mineral matter favors the production of HCN over NH3. Tsubouchi and co-workers51,71,102 also demonstrated that the Fe and Ca cations, which are inherently present in coals or through doping of demineralized coals, alter nitrogen distribution during pyrolysis. An analysis of results presented by Tsubouchi101,102 indicates that coals with high mineral matter content tend to release reduced HCN and increased N2. This analogy gives the reason behind the obtained low HCN and high N2 emanating from the three high mineral matter coals utilized in this study as illustrated in Figure 2. Both fluidized bed and drop-tube furnace reactors possessed high temperature zones which provided necessary conditions for the occurrence of secondary reactions. The release of NH3 from the FB was much greater than HCN under all FB temperature conditions as indicated by the NH3/HCN ratios illustrated in Figure 6a. However, more HCN than NH3 was released from the DTF as shown by the NH3/HCN ratios in Figure 6b. The difference between results from the FB and DTF indicates that formation of NH3 and HCN are to a great extent dependent upon coal properties and pyrolysis conditions. Leppälahti and Koljen23 as well as Li and Tan8 clearly showed that the ratio of HCN/NH3 released from model compounds varies to a large extent as a result H


Article

Energy & Fuels

Figure 7. Total of NH3 and HCN liberated during pyrolysis of the three coals in FB and DTF.

conditions and coal properties. HCN is primarily formed at the inception of pyrolysis. Nonetheless, the FB conditions enabled secondary pyrolysis to prevail, enhancing the hydrogenation of HCN to form NH3. FB conditions were characterized by relatively low temperature, long solid residence time, and large coal particle sizes. The conditions employed in DTF (high temperature, low solid residence time, and small coal particle sizes) were favorable for the production of HCN. NH3 and HCN released from either a vitrinite-rich or an inertinite-rich coal vary with temperature.

requiring more energy to break the bonds. The high hydrogen content and relatively low aromaticity of Matimba coal suggest that it has high aliphatic components,34 of which the energy at 1400 °C caused the driving off of significant quantities of volatiles incorporating nitrogen from the edge located N-5 and N-6. Nitrogen evolution was found to increase significantly with the pyrolysis temperature as illustrated in Figure 1. The analysis of respective chars, reported elsewhere,24,34 showed that the char nitrogen content diminished appreciably with temperature. Most of the nitrogen is evolved during the later stages of devolatilization as HCN; hence the yields of HCN were higher at 1400 °C.

■

■

CONCLUSION The influence of SA bituminous coal properties on the release of nitrogenous species during pyrolysis in a bench-scale fluidized bed at 740−980 °C and in a drop-tube furnace at 1000−1400 °C was evaluated. The pyrolysis N product distribution is dependent upon the coal properties and pyrolysis temperature. Char-N yields were predominant from FB pyrolysis, whereas coal-N was mostly converted to N2 during DTF pyrolysis. Vitrinite-rich Matimba coal released more volatile-N than the inertinite-rich Glisa and Lethabo coals. The fuel nitrogen in vitrinite-rich coal particles is released more easily than in inertinite-rich coals. However, there is no dominant coal property that can be singled out as being responsible for influencing the formation of a particular volatile-N product across the entire temperature range, but rather a combination or singular mineral matter and/or petrographic properties. Nonetheless, at a distinct point or specified temperature range, a particular coal attribute or combination determines the dominant volatile-N species liberated. Low temperature tars liberated from the three coals possessed similar nitrogen functional form attributes as the respective parent coals. The tars released at low temperatures (740−900 °C) predominantly consisted of pyrrolic nitrogen, followed by pyridinic and quaternary nitrogen, respectively. Relatively minute proportions of protonated and/or oxidized pyridinic nitrogen emerged in high temperature tars emanating from the vitrinite-rich coal. Tars liberated at the highest FB temperature (980 °C) mainly comprised quaternary nitrogen, followed by pyrrolic and pyridinic nitrogen, respectively. Mineral matter has a significant influence on the release of NH3 at the lowest FB temperature (740 °C). However, the release of NH3 at 820 °C correlated with the proportion of vitrinite maceral in the respective parent coals. Overall, higher yields of NH3 than HCN were obtained from the FB pyrolysis experiments; however, the opposite was true for DTF experiments as more HCN than NH3 was obtained. The formation of NH3 and HCN during pyrolysis of coal in FB and DTF is perceived to occur through different mechanisms due to different

AUTHOR INFORMATION

Corresponding Author

*E-mail: [email protected]. ORCID

Zebron Phiri: 0000-0002-6389-8490 Notes

The authors declare no competing financial interest.

■

ACKNOWLEDGMENTS The authors gratefully acknowledge the financial support of this research by the Eskom Power Plant Engineering Institute (EPPEI) and the Eskom Tertiary Education Support Program (TESP). The authors also express immeasurable gratitude to the Eskom Research and Innovation Centre (ERIC) and the Council for Scientific and Industrial Research (CSIR) for the permission to conduct experiments on the drop-tube furnace and the benchscale bubbling fluidised bed, respectively. Assistance rendered by Kganuwi Kekana, Buhle Metsing, and Brian Ncube on the DTF is greatly appreciated.

■

REFERENCES

(1) Johnsson, J. E. Fuel 1994, 73, 1398−1415. (2) Tsubouchi, N.; Ohtsuka, Y. Fuel Process. Technol. 2008, 89, 379− 390. (3) Tian, F.-J.; Wu, H.; Yu, J.-l.; McKenzie, L. J.; Konstantinidis, S.; Hayashi, J.-i.; Chiba, T.; Li, C.-Z. Fuel 2005, 84, 2102−2108. (4) Xie, Z.; Feng, J.; Zhao, W.; Xie, K. C.; Pratt, K.-C.; Li, C.-Z. Fuel 2001, 80, 2131−2138. (5) Leppälahti, J. Fuel 1995, 74, 1363−1368. (6) Tan, L. L.; Li, C.-Z. Fuel 2000, 79, 1883−1889. (7) Tan, L. L.; Li, C.-Z. Fuel 2000, 79, 1891−1897. (8) Li, C.-Z.; Tan, L. L. Fuel 2000, 79, 1899−1906. (9) Ledesma, E. B.; Li, C.-Z.; Nelson, P. F.; Mackie, J. C. Energy Fuels 1998, 12, 536−541. (10) Nelson, P. F.; Kelly, M. D.; Wornat, M. J. Fuel 1991, 70, 403−407. (11) Chang, L.-P.; Xie, Z.-L.; Xie, K.-C.; Pratt, K. C.; Hayashi, J.; Chiba, T.; Li, C.-Z. Fuel 2003, 82, 1159−1166.

I


Article

Energy & Fuels (12) Hämäläinen, J. P.; Aho, M. J.; Tummavuori, J. L. Fuel 1994, 73, 1894−1898. (13) Kidena, K.; Hirose, Y.; Aibara, S.; Murata, S.; Nomura, M. Energy Fuels 2000, 14, 184−189. (14) Carpenter, A. M.; Davidson, R. M.; Fukasawa, K.; Garner, L.; Mitchell, S. C.; Moreea-Taha, R.; Scott, D. H.; Sloss, L. L.; Smith, I. M.; Soud, H. N.; Takeshita, M.; Wu, Z. NOx Emissions and Control; IEA Clean Coal Centre: London, U.K., 2006; pp 1−299. (15) Solomon, P. R.; Colket, M. B. Fuel 1978, 57, 749−755. (16) Phong-Anant, D.; Wibberley, L. J.; Wall, T. F. Combust. Flame 1985, 62, 21−30. (17) Bassilakis, R.; Zhao, Y.; Solomon, P. R.; Serio, M. A. Energy Fuels 1993, 7, 710−720. (18) Genetti, D.; Fletcher, T. Energy Fuels 1999, 13, 1082−1091. (19) Baumann, H.; Möller, P. Erdöl Kohle-Erdgas Petrochem. 1991, 44, 29−33. (20) Chen, J. C.; Niksa, S. Symp. (Int.) Combust., [Proc.] 1992, 24, 1269−1276. (21) Mori, H.; Asami, K.; Ohtsuka, Y. Energy Fuels 1996, 10, 1022− 1027. (22) Glarborg, P.; Jensen, A. D.; Johnsson, J. E. Prog. Energy Combust. Sci. 2003, 29, 89−113. (23) Leppälahti, J.; Koljonen, T. Fuel Process. Technol. 1995, 43, 1−45. (24) Phiri, Z.; Everson, R. C.; Neomagus, H. W. J. P.; Wood, B. J. J. Anal. Appl. Pyrolysis 2017, 125, 127−135. (25) Kambara, S.; Takarada, T.; Yamamoto, Y.; Kato, K. Energy Fuels 1993, 7, 1013−1020. (26) Baxter, L. L.; Mitchell, R. E.; Fletcher, T. H.; Hurt, R. H. Energy Fuels 1996, 10, 188−196. (27) Haussmann, G. J.; Kruger, C. H. Symp. Combust., [Proc.] 1991, 23, 1265−1271. (28) Pohl, J. H.; Sarofim, A. F. Symp. (Int.) Combust., [Proc.] 1977, 16, 491−501. (29) Blair, D.; Wendt, J.; Bartok, W. Symp. (Int.) Combust., [Proc.] 1977, 16, 475−489. (30) Bruinsma, O. S. L.; Geertsma, R. S.; Oudhuis, A. B. J.; Kapteijn, F.; Moulijn, J. A. Fuel 1988, 67, 1190−1196. (31) Takagi, H.; Isoda, T.; Kusakabe, K.; Morooka, S. Energy Fuels 1999, 13, 934−940. (32) Freihaut, J. D.; Proscia, W. M.; Seery, D. J. Energy Fuels 1989, 3, 692. (33) Li, C.-Z.; Nelson, P. F.; Ledesma, E. B.; Mackie, J. C. Symp. Combust., [Proc.] 1996, 26, 3205−3211. (34) Phiri, Z.; Everson, R. C.; Neomagus, H. W. J. P.; Wood, B. J. Appl. Energy 2018, 216, 414−427. (35) Zhang, H.; Fletcher, T. H. Energy Fuels 2001, 15, 1512−1522. (36) Solomon, P. R.; Hamblen, D. G.; Carangelo, R. M.; Krause, J. L. Symp. (Int.) Combust., [Proc.] 1982, 19, 1139−1149. (37) Aho, M. J.; Hämäläinen, J. P.; Tummavuori, J. L. Combust. Flame 1993, 95, 22−30. (38) Niksa, S.; Cho, S. Energy Fuels 1996, 10, 463−473. (39) Rüdiger, H.; Greul, U.; Spliethoff, H.; Hein, K. R. Fuel 1997, 76, 201−205. (40) Friebel, J.; Köpsel, R. F. W. Fuel 1999, 78, 923−932. (41) Hämäläinen, J. P.; Aho, M. J. Fuel 1996, 75, 1377−1386. (42) Tian, F.-J.; Li, B.-Q.; Chen, Y.; Li, C.-Z. Fuel 2002, 81, 2203− 2208. (43) Li, X.; Zhang, S.; Yang, W.; Liu, Y.; Yang, H.; Chen, H. Energy Fuels 2015, 29, 7474−7482. (44) Deng, L.; Jin, X.; Zhang, Y.; Che, D. Energy Fuels 2013, 27, 430− 439. (45) Furimsky, E.; Nielsen, M.; Jurasek, P. Energy Fuels 1995, 9, 439− 447. (46) Pels, J. R.; Kapteijn, F.; Moulijn, J. A.; Zhu, Q.; Thomas, K. M. Carbon 1995, 33, 1641−1653. (47) Hansson, K.-M.; Åmand, L.-E.; Habermann, A.; Winter, F. Fuel 2003, 82, 653−660. (48) Schmiers, H.; Friebel, J.; Streubel, P.; Hesse, R.; Köpsel, R. Carbon 1999, 37, 1965−1978.

(49) Nielsen, M.; Jurasek, P.; Hayashi, J.; Furimsky, E. J. Anal. Appl. Pyrolysis 1995, 35, 43−51. (50) Leichtnam, J.-N.; Schwartz, D.; Gadiou, R. J. Anal. Appl. Pyrolysis 2000, 55, 255−268. (51) Tsubouchi, N.; Ohtsuka, Y. Fuel 2002, 81, 1423−1431. (52) Schäfer, S.; Bonn, B. Fuel 2000, 79, 1239−1246. (53) Schäfer, S.; Bonn, B. Fuel 2002, 81, 1641−1646. (54) Kambara, S.; Takarada, T.; Toyoshima, M.; Kato, K. Fuel 1995, 74, 1247−1253. (55) Stańczyk, K. Energy Fuels 1999, 13, 82−87. (56) Jones, J. M.; Zhu, Q.; Thomas, K. M. Carbon 1999, 37, 1123− 1131. (57) Li, C.-Z.; Nelson, P. F. Energy Fuels 1996, 10, 1083−1090. (58) Yuan, S.; Zhou, Z.-j.; Li, J.; Wang, F.-c. Appl. Energy 2012, 92, 854−859. (59) Espinal, J. F.; Truong, T. N.; Mondragón, F. Carbon 2007, 45, 2273−2279. (60) Kelemen, S. R.; Gorbaty, M. L.; Kwiatek, P. J.; Fletcher, T. H.; Watt, M.; Solum, M. S.; Pugmire, R. Energy Fuels 1998, 12, 159−173. (61) Hoogendoorn, A.; Rozendaal, C. M.; Bennecker, R. J. M.; Boersma, D. Behaviour of nitrogen during devolatilization of coal in relation to NOx pollutant emissions; KEMA Report; KEMA Laboratories: Arnhem, Netherlands, 1995. (62) Jacobson, I. A.; Heady, H. H.; Dinneen, G. V. J. Phys. Chem. 1958, 62, 1563−1565. (63) Patterson, J.; Drenchko, P. J. Org. Chem. 1962, 27, 1650−1650. (64) Nelson, P. F.; Li, C.-Z.; Ledesma, E. Energy Fuels 1996, 10, 264− 265. (65) Falcon, R.; Ham, A. J. J. South. Afr. Inst. Min. Metall. 1988, 88, 145−161. (66) Cai, H.-Y.; Kandiyoti, R. Energy Fuels 1995, 9, 956−961. (67) Cadle, A. B.; Cairncross, B.; Christie, A. D. M.; Roberts, D. L. Int. J. Coal Geol. 1993, 23, 117−157. (68) Hayashi, J.-i.; Kusakabe, K.; Morooka, S.; Nielsen, M.; Furimsky, E. Energy Fuels 1995, 9, 1028−1034. (69) Ohtsuka, Y.; Zhiheng, W.; Furimsky, E. Fuel 1997, 76, 1361− 1367. (70) Tsubouchi, N.; Ohtsuka, Y. Fuel 2002, 81, 2335−2342. (71) Tsubouchi, N.; Ohshima, Y.; Xu, C.; Ohtsuka, Y. Energy Fuels 2001, 15, 158−162. (72) Wu, Z.; Sugimoto, Y.; Kawashima, H. Energy Fuels 2000, 14, 1119−1120. (73) Wu, Z.; Sugimoto, Y.; Kawashima, H. Fuel 2003, 82, 2057−2064. The fourth UK Meeting on Coal Research and its Applications. (74) Hüttinger, K. J.; Michenfelder, A. W. Fuel 1987, 66, 1164−1165. (75) Yan, X.; Che, D.; Xu, T. Fuel Process. Technol. 2005, 86, 739−756. (76) Tyler, R. J.; Schafer, H. N. Fuel 1980, 59, 487−494. (77) Wu, Z.; Ohtsuka, Y. Energy Fuels 1997, 11, 477−482. (78) Ohtsuka, Y.; Watanabe, T.; Asami, K.; Mori, H. Energy Fuels 1998, 12, 1356−1362. (79) Ohtsuka, Y.; Mori, H.; Watanabe, T.; Asami, K. Fuel 1994, 73, 1093−1097. Coal Utilization and the Environment. (80) Roberts, M. J.; Everson, R. C.; Neomagus, H. W. J. P.; Okolo, G. N.; Van Niekerk, D.; Mathews, J. P. Fuel 2015, 158, 591−601. (81) Zhao, Y.; Hu, H.; Jin, L.; He, X.; Wu, B. Fuel Process. Technol. 2011, 92, 780−786. (82) Given, P. H.; Spackman, W.; Davis, A.; Zoeller, J.; Jenkins, R.; khan, R. Fuel 1984, 63, 1655−1659. (83) Rajan, S.; Raghavan, J. In Coal Science; Pajares, J., Tascón, J., Eds.; Coal Science and Technology; Elsevier: Amsterdam, 1995; Vol. 24, pp 1815−1818. (84) Everson, R. C.; Neomagus, H. W. J. P.; Kaitano, R.; Falcon, R.; van Alphen, C.; du Cann, V. M. Fuel 2008, 87, 3082−3090. (85) ISO. ISO 11760:2005. Classification of coals; Geneva, Switzerland, 2005. (86) ASTM International. ASTM D4326-04. Standard Test Method for Major and Minor Elements in Coal and Coke Ash By X-Ray Fluorescence; West Conshohocken, PA, 2004. J


Article

Energy & Fuels (87) ISO. ISO 1171:2010. Solid mineral fuels − Determination of ash; Geneva, Switzerland, 2010. (88) Speight, J. G. The Chemistry and Technology of Coal, 3rd ed.; Chemical Industries; CRC Press: Boca Raton, FL, 2012. (89) ISO. ISO 562:2010. Hard coal and coke − Determination of volatile matter; Geneva, Switzerland, 2010. (90) ISO. ISO 29541:2010. Solid mineral fuels − Determination of total carbon, hydrogen and nitrogen content − Instrumental method; Geneva, Switzerland, 2010. (91) ISO. ISO 7404:2. Methods for the petrographic analysis of bituminous coal and anthracite − Part 2: Preparation of coal samples; Geneva, Switzerland, 1985. (92) Nelson, P. F.; Smith, I. W.; Tyler, R. J.; Mackie, J. C. Energy Fuels 1988, 2, 391−400. (93) Xu, W.-C.; Kumagai, M. Fuel 2002, 81, 2325−2334. (94) EPA. U.S. EPA OTM: 29. Sampling and Analysis for Hydrogen Cyanide Emissions from Stationary Sources; 2011. https://www3.epa.gov/ ttnemc01/prelim/otm29.pdf. (95) EPA. US EPA CTM: 27. Procedure for Collection and Analysis of Ammonia in Stationary Sources; 1997. https://www3.epa.gov/ ttnemc01/ctm/ctm-027.pdf. (96) Li, C.-Z.; Buckley, A. N.; Nelson, P. F. Fuel 1998, 77, 157−164. (97) Nelson, P. F.; Kelly, M. D. In Proceedings of the 7th International Conference on Coal Science, Banff, Alberta, Canada, Sept 12−17, 1993; pp 140−143. (98) Wu, Z.; Ohtsuka, Y. Energy Fuels 1996, 10, 1280−1281. (99) Kelemen, S. R.; Afeworki, M.; Gorbaty, M. L.; Kwiatek, P. J.; Solum, M. S.; Hu, J. Z.; Pugmire, R. J. Energy Fuels 2002, 16, 1507−1515. (100) Stańczyk, K.; Dziembaj, R.; Piwowars, Z.; Witkowski, S. Carbon 1995, 33, 1383−1392. (101) Tsubouchi, N. Energy Fuels 2014, 28, 5721−5728. (102) Tsubouchi, N.; Abe, M.; Xu, C.; Ohtsuka, Y. Energy Fuels 2003, 17, 940−945.

K


Release of Nitrogenous Volatile Species from South African

Recommend Documents